PROJECT SUMMARY Traumatic brain injury (TBI) is a leading cause of death in the US, and treatment options are limited. Therapeutic clinical trials in TBI have yielded disappointing results owing in part to the difficulty in accounting for clinically important heterogeneity within TBI. Early delivery of therapy is essential after TBI to reduce secondary brain injury, but unrestricted treatment of all brain injuries could be harmful. TBI stimulates a complex cascade of immunologic responses, both centrally and peripherally. These peripheral immune responses to TBI could serve as an early sensor of risk phenotype given the rapid, readily measurable response in the blood. An improved ability to risk-stratify patients on admission will streamline patient selection for aggressive interventions?such as invasive neuromonitoring?versus selection of those patients who can safely be observed reducing potential harms. Holly E Hinson, MD MCR is a Neurologist and Neurointensivist at Oregon Health and Science University where she cares for patients with severe acute brain injury. The objective of this application is to develop supervised learning models of actionable short- and long-term outcomes post-TBI and to interrogate if pre-specified immunoregulatory proteins add predictive power to the models over clinical features alone. Her central hypothesis is that immunoregulatory proteomic signatures improve our ability to classify a low-risk clinical phenotype after TBI. Dr. Hinson?s preliminary data suggest peripheral cytokine levels are associated with actionable clinical events acutely after TBI. The project employs a highly-sensitive, single molecule immunoarray (SIMOA) to detect immunoregulatory proteins complemented with an unbiased proteomic approach utilizing global discovery mass spectrometry. She will develop and assess a series of models incorporating proteomic signatures to classify: acute progressive intracranial hemorrhage (Aim 1A), acute neurologic deterioration (Aim 1B), and long-term outcomes measured by the 6-month Glasgow Outcome Scale (Aim 2). She will develop these models in a well-defined, clinical trial population (development set), and test their ability to correctly classify outcome in an independent, prospectively enrolled cohort at OHSU (test set). Under a multidisciplinary team of expert mentors, the project will generate new insights into low-risk phenotype recognition and outcome classification after acute TBI. The proposed patient-oriented research project will be enhanced by a structured didactic program in the principles of predictive modeling and patient phenotyping (including proteomics), which will provide Dr. Hinson with the critical skills she will need to conduct independent, innovative translational clinical research in the field of neurotrauma.